As China grapples with intensified storms and flooding over recent months, authorities face an unexpected secondary crisis: a surge of AI-generated fake videos flooding social media platforms. These manipulated clips, ranging from fabricated rescue operations to false claims of casualties, are causing real-world panic and complicating emergency response efforts.
The misinformation wave includes highly realistic but entirely synthetic content. Videos have depicted bodies floating in floodwaters, crocodiles escaping into rivers, and collapsed infrastructure in areas untouched by disasters. Some creators edit overseas or historical footage, stripping watermarks and altering subtitles to present old events as current emergencies. Others use generative AI tools to produce convincing scenes of overflowing dams, emergency vehicles, and even fictional witness interviews.
This deluge has tangible consequences. In certain regions, false videos claiming imminent power outages triggered panic buying of emergency supplies. Misleading content about rescue operations has undermined public trust in official responses. The Chinese government has identified specific cases, such as a flood rescue video and footage of "released crocodiles," as examples of AI-generated material deliberately misrepresenting the situation to mislead the public.
Government crackdown and enforcement
In response, Chinese authorities launched a nationwide campaign on July 23 to remove illegal and harmful disaster-related online content. The Cyberspace Administration of China and the Ministry of Emergency Management issued a joint notice targeting misleading footage, recycled reports, fabricated data, fake official announcements, and AI-generated material. Police have arrested and penalized numerous individuals, with punishments ranging from detention to fines. Under the "Clean Net 2026" campaign, the Ministry of Public Security's Cybersecurity Bureau detailed 20 cases involving AI tools used to create fake videos, repackaged flood footage from other regions, and fabricated disaster claims to attract online traffic.
One notable case involved a 45-year-old man from Chengde, Hebei province, who created an AI-generated video claiming heavy rain caused ground collapse in Kuancheng county, sending vehicles into water. Posted on WeChat to gain followers, the video triggered widespread online discussion, caused local panic, and disrupted flood prevention efforts.
Experts attribute the problem to dramatic advances in generative artificial intelligence. Modern AI systems can now generate highly convincing disaster scenes, imitate news broadcasts, and produce realistic content with minimal technical skill. As Professor Chen Bing noted in an article for the Central Party School's Study Times, the barrier to creating rumors has significantly lowered: content generators need only capture hot topics and use AI tools to rapidly produce text, images, audio, and video, with even "one-sentence commands" yielding deceptive information that fuels mass rumor production.
Social media platforms face pressure to strengthen content review mechanisms. Authorities urge the public to approach disaster-related information from unknown sources with skepticism and rely on official announcements. Police emphasize trusting accredited sources during emergencies, though identifying AI-generated content remains increasingly difficult. As Typhoon Maysak and other recent weather events continue to affect regions like Guangxi province—where at least four died and 62,000 were evacuated—the battle against misinformation has become as critical as the physical disaster response itself.
Read the original article: